Revisiting the Myth: Husserl and Sellars on the GivenAuthor(s): Gail SofferSource: The Review of Metaphysics, Vol. 57, No. 2 (Dec., 2003), pp. 301-337Published by: Philosophy Education Society Inc.Stable URL: http://www.jstor.org/stable/20131977Accessed: 03/01/2010 12:11

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=pes.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

Philosophy Education Society Inc. is collaborating with JSTOR to digitize, preserve and extend access to TheReview of Metaphysics.

http://www.jstor.org

http://www.jstor.org/stable/20131977?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=pes

REVISITING THE MYTH: HUSSERL AND SELLARS ON THE GIVEN

GAIL SOFFER

I

In Science, Perception, and Reality, Sellars marvels at the power

of fashion in philosophy, which all too often offers us the spectacle of

a stampede rather than a careful sifting of gold from dross.1 Sellars

was worried that the flight from phenomenalism would lead to the fa

miliar pendulum effect and so thwart his effort to "usher analytic phi

losophy out of its Humean and into its Kantian stage," as Rorty has put

it.2 Accordingly, Sellars's critique of the Myth of the Given aimed to

show that what was really wrong with phenomenalism was nothing

particular to the sense-data of the positivists but the framework of

givenness itself.

Sellars's critique of the Given made an enormous impact, but did

it not set off a new stampede? On the contemporary horizon, the dust

is settling around numerous philosophies, including phenomenology,

and the hoofprints look strikingly like those of the postpositivistic an

alytic philosophers who are Sellars's progeny. And, some might think,

not without good reason. Are not phenomenology and phenomenal

ism kindred cousins? Does not the notion of an immediate, private

"given" lie so much at the core of the Husserlian enterprise that phe

nomenology is unthinkable without it? Does not Husserl wish to erect

the edifice of objective knowledge on this unshakable foundation?

Does not Sellars extend his critique of the Given to Chisholm and so to

Brentano and Husserl?3 When we come to the essence of the matter,

Correspondence to: 549 West 123d Street -

17B, New York, NY 10027. 1 Wilfrid Sellars, "Phenomenalism," in Science, Perception and Reality

(Atascadero: Ridgeview Publishing Company, 1991), 60. 2 Richard Rorty, "Introduction" to Wilfrid Sellars, Empiricism and the

Philosophy of Mind (Cambridge: Harvard University Press, 1997), 3. 3 Wilfrid Sellars and Roderick Chisholm, "Intentionality and the Mental,"

in Minnesota Studies in the Philosophy of Science, ed. Herbert Feigl, Michael Scriven, and Grover Maxwell, vol. 2 (Minneapolis: University of Min nesota Press, 1958), 527.

The Review of Metaphysics 57 (December 2003): 301-337. Copyright ? 2003 by The Review of

Metaphysics

302 GAIL SOFFER

can we not treat Husserl and Russell in one breath as an indivisible

unity, as does Rorty throughout Philosophy and the Mirror of Na

ture?4 In the wake of Sellars, Wittgenstein, Rorty, the linguistic turn, and the rest, would we not do well only to mention and not to use the

word "given" in polite philosophical company?

According to the spirit if not the letter of Sellars, it is my hope in this paper to redimension the relationship between postpositivistic

analytic philosophy and Husserlian phenomenology by revisiting some of the arguments advanced by Sellars against the Myth of the

Given and in favor of the linguistic turn. One thesis advanced in this

paper is that there is more common ground between Husserl and Sell

ars than is usually thought. For the main aim of Sellars's critique is to

attack the given as the immediate and to show that empirical knowl

edge requires concepts, inferences, and language. Here there is little

if any disagreement with Husserl, who is hardly concerned with im

mediacy and makes very similar points about the mediacy of empiri

cal knowledge, albeit in a different way.

However, certain divergences remain, and a second aim of this

paper is to reconstruct and evaluate them. One important dispute concerns the relation between language and intentionality. For

Sellars and his progeny, language is a precondition for attentive, ob

ject-directed consciousness; whereas on a phenomenological ac

count, there are prepredicative, preverbal forms of intentionality. A

second, related difference arises in the respective approaches to inter

subjectivity. Sellars treats the attribution of intentionality to others as

primarily a theory for the explanation of behavior; whereas for Hus

serl the "constitution" of the other is not primarily theoretical, and a

relation to behavior does not belong to the very concept of intention

ality.

Behind these disputes lie certain deeper differences concerning

the nature of the mental and the relation between science and every

day life. For there is a secondary tendency in Sellars's linguistic turn

which shies away from the given not as the immediate but more gen

erally as the subjective, experiential dimension of life, which Sellars

sometimes associates with the Cartesian, "mind's eye" view of con

sciousness. Sellars's functional-linguistic approach to intentionality

aims to define the mental without making any explicit reference to the

4 Richard Rorty, Philosophy and the Mirror of Nature (Princeton: Prin ceton University Press, 1979), 4, 166-9, 369, 390.

HUSSERL AND SELLARS ON THE GIVEN 303

properly subjective dimension of mental life. This approach is meant

to clear the way for an eventual reduction of the mental to the physi cal in neurophysiological terms. By contrast, according to phenome

nology as well as antireductionists such as Nagel, Searle, and to some

extent Chalmers, the experiential dimension is essential to conscious

ness. The experiential dimension cannot be eliminated, translated, or

reduced, and any conception which lacks explicit reference to it is not

a conception of consciousness but of its physical causes, its functional

relations, or something else.

Sellars is motivated to accommodate his conception of the men

tal to physicalism by a belief that the manifest and the scientific im

ages are ultimately incompatible. He argues that if we were to believe

both the common sense and scientific descriptions of the world, then

as in Eddington's example, there would be two tables instead of one

(that is, the table as described by common sense and the table as de

scribed by mathematical physics). Sellars concludes that since there

is only one world, only one image can be true, and this is the scientific

image. Much as the secondary properties of the manifest image were

replaced by purely mechanical properties in seventeenth-century

physics, so the perceptions, thoughts, and feelings of the common

sense experience of the world will be replaced by neurophysiological states in a more advanced scientific image.5

However, here I think Sellars is in error. As I have argued in de

tail elsewhere, Husserlian phenomenology can be used effectively to

show that there is no either-or between the mathematized world of na

ture and the lifeworld: both the lifeworld and mathematized nature are

real, albeit in different ways and according to different views of the

real.6 There are not two tables but one "object" which is conceptual ized in different ways, according to different frameworks, language

games, and ontological attitudes. The scientific, naturalistic concep

tion of the world can neither claim universal, absolute applicability nor invalidate the notion of the real appropriate to the manifest image

(the natural attititude).

Thus, the general upshot of my analysis in this paper is that there

is little or nothing in Sellars, the critique of the Given, or the linguistic

5 Wilfrid Sellars, "Philosophy and the Scientific Image of Man," Science, Perception and Reality (Atascadero: Ridgeview, 1963). 6 For a discussion of ontological pluralism, see Gail Soffer, "Phenome

nology and Scientific Realism: Husserl's Critique of Galileo," Review of Meta

physics 44, no. 1 (September 1990): 67-94.

304 GAIL SOFFER

turn that could justify the superficial Rortean assessment of Husser

lian phenomenology as "superseded" by contemporary philosophical

developments. To the contrary, postpostivistic analytic philosophy

has much to learn from Husserl, especially in the areas of the relation

between language and thought and of intersubjectivity.

II

In mounting an Auseinandersetzung between Husserl and Sell

ars, we need to guard against certain facile interpretations, such as

that Husserl believes there is a given, or that Husserl has a Cartesian

view of the mind, whereas Sellars does not.7 That it is not a simple

matter to pinpoint the differences is clear in Sellars's correspondence

with Chisholm, whose view of intentionality is relatively similar to

Husserl's. In this correspondence, we find a constant nuancing of dif

ferences, a veritable Sisyphean striving toward an agreement just out

of reach. Even Chisholm's blunt statement of his position in seven

points, including the crucial sixth point, "thoughts are a 'source of in

tentionality'," does not elicit an outright rejection from Sellars but a

hedging acceptance: "it isn't so much that I disagree with your seven

sentences, for I can use each of them separately, with varying degrees

of discomfort, to say something which needs to be said.... It is rather

that I am unhappy about the force they acquire in the over-all frame

work in which you put them."8

The initial obstacle to evaluating how phenomenology might fall

into the "framework of givenness" is that Husserl and Sellars have

fundamentally different conceptions of givenness itself. To see how

much of the hostility to phenomenology on the part of Sellars's prog

eny is based on mere equivocation, we must clarify how each thinker

understands the term "given."

For Sellars, the given is the immediate in the sense of the un

learned. It is the content of an awareness which does not require

prior experience or training to be grasped but is simply "there," in the

7 According to Sellars, no philosopher would object to talk of the "given"

in the sense of a distinction between what is perceived and what is inferred. See Sellars, Empiricism and the Philosophy of Mind, 13. Further, in his cor

respondence with Chisholm, Sellars grants that thought is distinct from lan

guage and can occur in its absence. See Sellars and Chisholm, "Intentionality and the Mental," 527-8.

8 Sellars and Chisholm, "Intentionality and the Mental," 535.

HUSSERL AND SELLARS ON THE GIVEN 305

raw. It is a type of awareness which does not presuppose language,9

concepts, or inferences. Sellars does not deny the existence of imme

diate contents of this form. Rather, what he opposes is the idea that

this type of awareness can serve as evidence for empirical knowledge.

For Sellars, the essence of the Myth of the Given is to think there is

nonlinguistic, nonconceptual, noninferential awareness which either

serves as evidence for or itself constitutes empirical knowledge. In

deed, according to Sellars, the whole point of the empiricists' inven

tion of the epistemological category of the given is to identify the type

of awareness which: (1) does not require learning, concepts, or lan

guage; and (2) whose occurrence logically entails the existence of em

pirical knowledge.10 Now it does not require a deep familiarity with Husserl to realize

that this is not the point of the category of givenness for Husserl, and

that he, like Sellars, denies there is anything to which both (1) and (2) apply. However, before coming to this in detail, it will be useful first

to outline the nature and function of givenness for Husserl.

Although "given" and its cognates occur with remarkable fre

quency in Husserl's writings, its ubiquity does not express a unitary

conception of "the" given. We can differentiate at least three main

senses of givenness in Husserl, none of which corresponds directly to

Sellars's notion of the immediate as the unlearned, the preconceptual,

preinferential, and prelinguistic. (Indeed, for Husserl, conceptuality,

linguisticality, and inferentiality are three separable features of

9 Here and throughout this paper, I follow Sellars's usage and employ "language" and its cognates to refer to specifically verbal behavior. Sellars does not present a general definition of language, but he takes speech as the central model, and seems to presuppose that language is: (1) propositional; (2) associated with a form of observable behavior; (3) something that adult humans exhibit but animals and infants lack. Thus his conception of lan

guage excludes developmental^ prior and passive communicative forms such as gestures, facial expressions, nonverbal vocalizations, and passive comprehension of speech. If Sellars had conceived of language more along the lines of Merleau-Ponty as (active and passive) corporeal intersubjective communication, and had held that the framework of intentionality presup posed language in this sense, then many of the reservations I will raise in what follows would have to be significantly qualified. But then the position of psychological nominalism would be about the fundamental role of inter

subjective communication, not specifically of words. 10

"[T]he point of the epistemological category of the given is, presum ably, to explicate the idea that empirical knowledge rests on a 'foundation' of non-inferential knowledge of matter of fact"; Sellars, Empiricism and the

Philosophy of Mind, 15.

306 GAIL SOFFER

consciousness and should not be lumped together under a single

heading.) In the broadest sense, the given for Husserl is what is experi

enced, as it is experienced, even where this experience presupposes

concepts, language, or inferences. Husserl uses this very general sense of givenness to emphasize that phenomenology is concerned

with a descriptive analysis of experience, as opposed to, for example, a hypothetico-deductive explanation of its physical causes.

The second, narrower sense of givenness is best captured by Husserl's technical notion of immanence. A content or a process is

immanent if and only if it cannot be thought as existing apart from

consciousness. In Husserlian terminology, the immanent is a non

independent part or abstract moment of consciousness itself.11 For

example, a table that exists in the world is not immanent, but an Ab

schattung of a table is, and so are a memory of the table, the blurry

image I see when I press on my eye while looking at the table, and so

forth.

Immanence is the sense of givenness in Husserl's notion of das

Wie des Gegebenheit (the subjective mode of givenness), which delin

eates the phenomenological research program of constitutive analy sis. According to Husserl, the world as we experience it, the world as

meaningful, does not exist "in itself but becomes constituted for us

through the involvement of subjects and communities of subjects. The task of phenomenology is to bring to reflective awareness the

specific subjective elements and processes (the immanent) required for experience of various types of objects. The notion of das Wie des

Gegebenheit emerges from the insight that experience exhibits a dual

structure: on the one hand, we have objects and their properties; on

the other, we have the strictly subjective components of experience

(the immanent) on whose basis experience of objects is possible. For

example, enduring three-dimensional objects are experienced on the

basis of fleeting, partial perspectival views (Abschattungen and their

syntheses). Similarly, we experience the past and an objective time

order on the basis of subjective apprehensions which take place in the

present (for example, memories, reports of others). In ordinary life, our attention goes right through the subjective modes of manifesta

tion and concentrates itself on things, so that the subjective side of

11 Edmund Husserl, Logical Investigations, trans. J. N. Findlay, vol. 1

(New York: Humanities Press, 1970), Third Investigation, ?2.

HUSSERL AND SELLARS ON THE GIVEN 307

experience vanishes from view. It would not be an exaggeration to

say that the crux of phenomenology is to thematize the largely hidden

role of subjectivity in experience and to give a precise account of the

way in which our experience of things as objective depends upon it. It

is for this reason that Husserl characterizes the epoch? as a turning of

one's regard from the objects of the world to the objects in their mode

of givenness (das Wie des Gegebenheit).12 A closely related view is ex

pressed in Heidegger's notion in Being and Time that phenomenology shifts attention from beings to their Being.

Here again, there is no correspondence between Husserl's con

ception of givenness as immanence and Sellars's notion of immediacy.

A content is immanent when it is an intrinsic part of consciousness, in

dependently of whether occurrence of the content presupposes con

cepts, language, learning, and so forth. For example, a table-Abschat

tung is immanent, but the ability to experience it is clearly acquired,

and, depending upon its epistemic sophistication, could require con

cepts and even language. A third sense of givenness goes under Husserl's term Selbstgeben

heit (self-givenness or original givenness) and is associated with his

analysis of Evidenz (self-evidence). As a first approach, we can corre

late original givenness with self-manifestation, in contrast to depiction

through resemblances or indication through symbols. An object is

originally given when it is perceived, as opposed to merely thought. For example, if I look out a window and see a blackbird flying by, the

flying blackbird is originally given. By contrast, if I judge "There is a

blackbird flying outside the window" without looking (for example, because someone told me so), the blackbird is not originally given.

The concept of original givenness would be banal if not for its rel

ative nature. One experience makes the object manifest in compari son to another experience, and there is a wide spectrum of degrees.13

At one extreme lies purely conceptual thought, trains of reflection

which are unaccompanied by images, memories, or perceptions of the

subject matter being contemplated. This type of empty representation is characterized by Husserl as symbolic since there is no resemblance

between the medium by which we are related to the matters under

consideration (for example, the words of one's thoughts) and the

12 Edmund Husserl, The Crisis of the European Sciences and Transcen dental Phenomenology, trans. David Carr (Evanston: Northwestern Univer

sity Press, 1970), 143-7. 13 Husserl, Logical Investigations, vol. 2, Sixth Investigation, ?23.

308 GAIL SOFFER

matters themselves (for example, states of affairs in the world).14 Further along the spectrum of originality we find pictures, images, and memories: here we are related to subject matters by means of

copies and resemblances rather than merely by symbols. Toward the

far end of the spectrum we find perception, which is characterized by an encounter with the thing itself, in the flesh, the original rather than

a picture, copy, or symbol of the thing.15 Of course Husserl's assertion

that in perception we encounter the thing "itself has nothing to do

with the idea that all (or any) perception is veridical or that percep

tion is unmediated by concepts or language, much less that in percep

tion we encounter the Kantian Ding an sich. It is simply a way of

characterizing the difference between what it is like to hear a piano sonata being played or to see a friend and what it is like to see a pic ture of these events or to imagine, think, or read about them. In per

ception, it "appears" that the thing itself stands before us, or in Hus

serlian terms, we judge that it does so.

Perception as well is not monolithic but makes objects manifest

in different ways and to different degrees. According to Husserl's

analysis, perceptual acts have "parts" in the sense that they contain

numerous explicit and implicit judgments about the object per

ceived.16 For example, when we see a table, we implicitly judge what

it would feel like if touched, how it would look from alternate loca

tions, how it would be perceived by other persons under similar con

ditions, whether its material would be durable under normal wear

and-tear, whether it is large enough to use for writing, and so forth. In

most perceptions, only some of the properties attributed to the object

have the character of self-givenness (in the sense outlined in the pre

vious paragraph). For example, when one looks at a hammer, one

judges that it would be resistant to the touch (for otherwise one

would experience that one is seeing a phantom rather than a solid ob

ject), and perhaps even that this is precisely the tool for the task at

hand. However, while the visual aspect of the hammer is originally

given by merely looking, its tactile quality and its suitability for ham

mering are not. These latter would be originally given in grasping the

hammer and hammering with it.

The merely partial originality of perception motivates Husserl's

limit idea of maximal original givenness, or adequate Evidenz, which

14 Husserl, Logical Investigations, vol. 2, Sixth Investigation, ?14a.

15 Ibid. 16

Ibid., ?12.

HUSSERL AND SELLARS ON THE GIVEN 309

is a perception in which all the properties attributed to the object are

self-given, or in Husserl's terminology, all the implicit judgments are

fulfilled. In adequate self-evidence, there is nothing judged or implied

about the object which is not also perceived. The object manifests it

self exactly and completely as it is judged.17 For Husserl, adequate

self-evidence is givenness in the strict sense.

Having completed our sketch of adequate self-evidence, we

should note that Husserl's use of this sense of givenness is completely

consistent with Sellars's attacks against the sense-data of the logical

empiricists. According to Husserl, empirical sense-data cannot be ad

equately self-evident.18 The Ayersian exemplars of givenness attacked

by Sellars, such as "I see a red surface" or "the tomato presents a red

sense-datum to me," contain numerous implicit judgments which re

main unfulfilled. The assertion "the tomato presents a red sense-da

tum" is generally accompanied by the implicit judgment that the to

mato is a three-dimensional physical object with an appropriately

colored back side (for example, red or green), a certain taste and tac

tile quality, and numerous other properties. We might try to formulate

a judgment without unfulfilled implicit assumptions by asserting,

"something, which appears to be a tomato, presents a red sense-da

tum," or "red here now." However, even the minimalist report "red

here now" is not adequately self-evident if "red" is taken with its usual

meaning, according to which red-experiences are intersubjectively

shared. When taken with its usual meaning, "red here how" includes

the implicit judgment that the red I am now seeing is the same color

you see when you say "red," an implicit judgment which is merely pos

ited, not fulfilled. Ultimately, any implicit reference to intersubjectiv

ity or existence in the world comports expectations of future percep

tions whose realizability is merely presupposed. For this reason,

Husserl concludes that even to approach adequate self-evidence, an

object or content has to be taken as it appears to me in the here-and

now apart from any positing of objective reality or intersubjectivity;

that is, as it appears under the phenomenological reduction.

Thus we see that Husserl's concept of givenness as adequate self

evidence has little to do with Sellars's notion of immediacy as the ab

sence of conceptuality or language. Quite the contrary, Husserl's anal

ysis makes much the same point generally associated with Sellars:

17 Ibid, ?37.

18 Edmund Husserl, General Introduction to a Pure Phenomenology, trans. Frederick Kersten (The Hague: Martinus Nijhoff, 1980), ?44.

310 GAIL SOFFER

even seemingly very simple perceptions involve quite complex im

plicit judgments, which in turn generally require learning, concepts,

and language.

Together with their different conceptions of givenness, there is a

significant divergence between Husserl and Sellars over the purpose

of the category of the given. For Sellars, the point is to found empiri

cal knowledge, to identify the noninferential bases for inferences. By

contrast, for Husserl the category of the given serves to thematize the

subjective elements of experience (the immanent) and to show how

what is taken by us to be knowledge presupposes and emerges out of

these subjective elements.19 This exploration reveals the origins and

presuppositions of empirical knowledge and clarifies the motives for

belief. But what we find in the case of empirical knowledge is that the

motives are always insufficient to yield certainty. No empirical truths

follow as consequences of this type of analysis of the given.20 Indeed,

the given as the immanent does not in general provide an inferential

basis for empirical knowledge but rather motivates it. For example,

one does not infer from a series of table-Abschattungen that there is a

table, the Abschattungen do not serve as evidence that there is a ta

ble, one cannot justify the claim "there is a table" by asserting "I have

a series of table-Abschattungen." Rather, the justifying evidence for

the claim that there is a table is that one sees the table. The function

of the Abschattungen is not to justify but to give rise to the perception

of the table by motivating it, based on subjective mechanisms of asso

ciation, among others.

Ill

In the previous section I argued that Husserl is not committed to

the Myth of the Given in its basic form. However, for Sellars the Myth

is a veritable hydra, turning up from classical empiricism to logical

19 Edmund Husserl, Erste Philosophie (1923/24), Erster Teil, ed. Rudolf

Boehm, in Husserliana, vol. 7 (The Hague: Martinus Nijhoff, 1956), 356. 20 "No objective truth, whether in the prescientific or in the scientific

sense, that is, no claim about objective being, ever enters into our scientific

sphere, whether as a premise or as a consequence (Keine objektive Wahr

heit, ob in vorwissenschaftlichem oder wissenschaftlichem Sinne, bzw. keine Feststellung f?r objektives Sein tritt je in unserem Kreis der Wissen

schaftlichkeit, ob nun als Pr?misse oder also Folgerung)"', Husserl, Crisis

of the European Sciences, ?52.

HUSSERL AND SELLARS ON THE GIVEN 311

positivism, from Descartes to Hegel. In particular, Sellars holds that

rejecting the Myth of the Given means affirming that language is the

precondition for the most basic forms of cognition and perceptual awareness of objects. He defines his psychological nominalism as the

view that prior to language, there is no awareness of the logical space

of particulars, sorts, resemblances, facts, and so forth.21 Without lan

guage, one cannot even notice a certain sort of thing or property be

cause noticing something presupposes possessing the concept of that

thing, and this in turn presupposes a complex network of other con

cepts and social initiation into preexisting language games.

We [have to] give up the idea that we begin our sojourn in this world with any?even a vague, fragmentary, and undiscriminating?awareness of the logical space of particulars, kinds, facts, and resemblances. . . .

[T\nstead of coming to have a concept of something because we have noticed that sort ofthing, to have the ability to notice a sort ofthing is

already to have the concept of that sort of thing, and cannot account

for it.22

On this basis Sellars argues that the abstractive theory of concept formation of classical empiricism is another version of the Myth of the

Given.23 Concepts cannot be formed on the basis of perceptions of

particulars because without concepts one cannot see particulars, and

indeed properly speaking one does not see at all because seeing is

cognitive and cognition requires concepts and language.24 Husserl's account of concept formation differs in important ways

from the classical empiricist ones, as well as from the Thomistic-Aris

totelian notion of abstraction also attacked by Sellars. Nonetheless, Husserl shares the empiricist view that the noticing of individuals and

their individual features occurs prior to and as a condition of the for

mation of general concepts.

21 Sellars, Empiricism and the Philosophy of Mind, 66. 22 Ibid., 87.

23 "[T]he coming to see something as red is the culmination of a compli

cated process which is the slow building up of a multi-dimensional pattern of

linguistic responses (by verbal expressions, by meta-linguistic expressions to

object-language expressions, etc.) the fruition of which as conceptual occurs when all these dimensions come into play in such direct perceptions as that this physical object (not that one) over here (not over there) is (rather than

was) red (not orange, yellow, etc.)"; Sellars, "Phenomenalism," in Science, Perception and Reality, 90. 24

"Seeing is a cognitive episode that involves the framework of

thoughts"; Sellars, Empiricism and the Philosophy of Mind, 110.

312 GAIL SOFFER

However, here again we need to be careful. For whether there re

ally is a clash depends upon how one interprets Sellars's phrases,

"perceiving something as red" and "concept of red." In contrast to

Sellars and those under the influence of his linguistic turn, Husserl

painstakingly differentiates among many different degrees of

epistemic complexity of perceptions and concepts. Seen in light of

such distinctions, phrases such as "perceiving something as red" and

"concept of red" are highly ambiguous and could refer to a wide range

of phenomena, many of which would indeed presuppose verbal lan

guage and/or concepts, even on a Husserlian account. For example,

"perceiving something as red" could refer to a perceptual judgment

where:

(1) an individual color instance is subsumed under a general color con

cept (that is, "this instantiates the general color concept, red");

(2) an individual object which has already been subsumed under a gen eral concept then receives a predication of an individual color ("this ap ple is that");

(3) the individual color of an individual object which has already been subsumed under a general concept is now itself subsumed under a gen eral color concept ("this apple is red");

(4) a perceptual judgment where an individual color is predicated of an individual object ("this is that");

(5) a perceptual judgment in which the colored object stands out from the perceptual background but there is no separate noticing or predica tion of its color ("this-such");

(6) a perceptual judgment in which the colored object is experienced as

part of the perceptual background (for example, peripherally, sublimi

nally) but is not thematically noticed.

(7) perceptual judgments in which the individual color is associated with a previously seen individual colors or individual objects (as simi

lar, as identical, as different, as related), and this with or without the matic attention.

(8) the same as (7), only involving subsumption of some or all of the in dividuals under general concepts.

Similarly, the epistemic complexity of concepts varies greatly,

also as a function of social-linguistic interaction. The fact that verbal

linguistic interactions affect concepts as soon as these interactions

begin implies neither that there are no concepts prior to language nor

that concept formation logically presupposes verbal interactions.

HUSSERL AND SELLARS ON THE GIVEN 313

While in general Sellars adopts an overly simplistic all-or-nothing

notion of conceptuality, upon occasion he admits that there is a dis

tinction between rudimentary and sophisticated concepts. For exam

ple, in Empiricism and the Philosophy of Mind, Sellars imagines a tie

salesman, John, with two different types of color concepts. John be

gins with simple color concepts lacking any reference to standard

conditions or standard observers. At first John does not distinguish

between being and appearing and even thinks a green tie is blue when

it looks blue under artificial lighting. Through imitation and selective

verbal reinforcement, John learns that colors do not always look like

what they are and that they look like what they are when they are in

natural daylight, viewed by eyes in good condition, and so forth. That

is, he learns what type of conditions count as standard conditions and

what type of observers count as standard observers for the viewing of

colors. In this way, he arrives at more sophisticated color concepts. If

bn is John's na?ve concept of blue, and bs his sophisticated blue con

cept, we can say that something is bs if and only if it looks bn to a stan

dard observer in standard conditions.

Now Sellars uses the tie salesman example to argue that noticing

shades of blue requires a vast network of concepts and hence a so

cially reinforced language game. Yet in fact the example of the tie

salesman shows only that bs (the concept involving notions of stan

dard observers and conditions) presupposes language and a network

of other concepts, including bn. The example does nothing at all to

show that bn (the initial, primitive concept that equates being and ap

pearing) requires language or that noticings of individual blue-shades

require any concept of blue, even a rudimentary one.25 Thus, although Sellars here distinguishes between different degrees of conceptual so

phistication, he fails to realize the implications of this distinction for

his position.

It is clear that some concepts, and virtually all concepts employed

by normal adults in everyday life, are shaped in part by social rein

forcement, which occurs especially (although not solely) in the form

of verbal communication. For example, normal adult concepts of sen

sible qualities are at least as complex as John's sophisticated version

of the concept of blue and thus depend upon verbally reinforced no

tions of standard observers and standard conditions. Similarly to

25 For the argument involving the tie salesman, see Sellars, Empiricism and the Philosophy of Mind, 37-46.

314 GAIL SOFFER

Sellars, Husserl shows how a complex notion of normality is funda

mental to mature object-perception and that what counts as a normal

perception or observer in a given context evolves with intersubjective

interaction, and so with verbal communication.26 It is also clear that

"perceiving something as red" presupposes the concept of red for the

type of perceptions to which normal adults would usually refer with

this phrase (for example, (1), (3), and (8) above). Thus if when we

speak of perceiving something as red we have in mind a type of per

ception involving subsumption under a general concept, and if this

concept is understood by us to be of at least average adult sophistica

tion, then the perception of something as red presupposes the con

cept of red, a vast network of other concepts, and social initiation into

verbal language games.

However, this still leaves some senses of perceiving something as

red that do not obviously presuppose any concept of red. This would

include perceptions of individual reds or red things which are not

predicatively elaborated, with or without attention and association

(for example, (2), (4), (5), (6), and (7) above). It remains a debated in

terpretative issue whether for Husserl there is any sensible experi ence wholly free of conceptual order laid down by the intellect. Many

commentators believe that Husserl has a Kantian, antiempiricist view,

according to which even the genetically most primitive sensibility is

shaped in terms of certain basic conceptual distinctions, such as

sameness and difference, whole and part, feature and thing, presence

and absence, and so forth.27 But if a conceptual framework does oc

cur at the genetically most primitive level, it must be innate since the

results of social reinforcement would not belong to the genetically most primitive level. Thus, whatever his position on primal sensibil

ity, Husserl holds that there is some form of perception which is not

26 For an extensive discussion of the concept of normality in Husserl, see Anthony Steinbock, Home and Beyond (Evanston: Northwestern Univer

sity Press, 1995). 27 Husserl's term for the genetically most primitive level of experience is

"primal sensibility," sensibility prior to the operations of synthesis, associa

tion, learning, and so forth. His views on primal sensibility are provocative but sketchy and do not affect the argument advanced here. For a discussion of the possible penetration of concepts to the level of primal sensibility, see Robert Sokolowski, Presence and Absence (Bloomington: Indiana University Press, 1979); Iso Kern, Idee und Methode der Philosophie: Leitgedanken f?r eine Theorie der Vernunft (Berlin: Walter de Gruyter, 1976); and James Hart, "Agent Intellect and Primal Sensibility," in Issues in HusserVs Ideas II, ed.

Thomas Nenon and Lester Embree (Dordrecht: Kluwer, 1996).

HUSSERL AND SELLARS ON THE GIVEN 315

predicatively shaped and does not presuppose socially transmitted

verbal language. In the remainder of this section I will review Sellars's account of

prelinguistic consciousness and argue that it does not effectively sup

port his claim that there can be no prelinguistic intentionality.

Sellars's positive account of prelinguistic consciousness is contained

in his theory of sensory awareness. Acording to this, the conceptual

framework of sensations does not arise from first having sensations, then reflecting on them, then forming a concept of them by abstrac

tion. To the contrary, for Sellars sensations give a hypothetical expla nation of object perception. The hypothesis of sensations serves espe

cially to explain illusions, as when people report that they see a green

object while they are looking at a blue one. Using the sensation hy

pothesis, we say that under normal conditions a blue object causes a

blue sensation and a green object a green sensation, but under certain

unusual lighting a blue object causes a green sensation. More pre

cisely, the theory of sensations asserts that:

(1) a sensation is a state of the subject and not a relation of the subject to something else. Thus to sense is not to perceive a particular (that is, the sense impression itself) but to be in a certain state;28

(2) sensations are caused by physical objects impinging on the body;

(3) a sensation is a theoretical entity conceived on analogy with the ex ternal object which in normal circumstances is its physical cause;

(4) sensations are not conceived as possessing the same intrinsic prop erties as the physical objects that normally cause them. In Sellars's ter

minology, sensations have properties which are "formally analogous" (which correspond in their similarities and differences) to the similari ties and differences of the properties of their normal objective causes. The properties of sensations and the properties of their normal objec tive causes have analogous logical spaces. If ?h (j)2,... (j)n are properties of physical things, then the sense impressions normally caused by phys ical objects with these properties possess another set of properties ipi, ip2, ^n such that op!, %,

. . . ipn resemble and differ from one another

in a way formally analogous to the way h ?2,

316 GAIL SOFFER

explanation of the perception of objects in linguistically mature

adults; and as an account of the experience of conscious beings lack

ing verbal language. Now in the case of object perception, the theory of sensation is an empirical theory: there are certain observational

data to be explained (for example, illusions), a hypothesis is mounted

in terms of other existing theories (for example, causal theories of

perception, physical reductionism), the theory is confirmed by predic

tive-explanatory adequacy, and so on. However, in the case of non

verbal experience, the theory of sensation is not an empirical account.

Here the empirical data (such as the behavior of preverbal infants and

animals, the neurophysiology of prelinguistic beings) play no essential

role in the mounting of the theory. Instead, the theory that prelinguis tic consciousness consists of sensation is a philosophical position re

sulting from Sellars's psychological nominalism, according to which

language is the condition for concepts, awareness of particulars,

kinds, and so forth. Since prelinguistic beings have bodies but not

verbal language, and since according to psychological nominalism

there cannot be perception without verbal language, it follows, as it

were by default, that there can only be sensations, the same physically caused states that in linguistic beings result in object perception. The

burden of just-ification for the theory of sensation as an account of

prelinguistic consciousness thus falls squarely on Sellars's arguments for psycho-logical nominalism. I have suggested above that in fact

Sellars's arguments for psychological nominalism do not show that

there are no concepts or perceptions without language but only that

there cannot be concepts and perceptions of a certain epistemic so

phistication without language. If this is the case, then the brute equa

tion of prelinguistic consciousness with Sellarsian sensation is also

unjustified.

A related difficulty stemming from this philosophical derivation is that the theory can come to be, and in fact is, in contradiction with

prevailing empirical theories. Sellars works within the context of be

haviorism and shares its presupposition that noncognitive stimulus

response mechanisms can successfully explain even the most sophis

ticated behavior and learning exhibited by infants and animals.29

However, today most empirical researchers find that behaviorism is

not a successful research program and that many preverbal behav

ioral phenomena (including language acquisition itself) are much bet

29 Sellars and Chisholm, "Intentionality and the Mental," 527-8.

HUSSERL AND SELLARS ON THE GIVEN 317

ter explained and predicted by attributing perceptions, recognition,

rudimentary inferential processes, desires, and even beliefs to infants

and animals.30 Now according to Sellars, to have good reason for es

pousing a theory is to have good reason for saying that the entities it

postulates really exist.31 Since superior explanatory and predictive

power are as good reasons as any for espousing a theory, then follow

ing Sellars we should say that preverbal children and animals really have intentionality.32

A more fundamental question concerns whether Sellarsian sensa

tions, as in the case of all functionally defined states, are rightly char

acterized as states of "consciousness." Following Nagel and Searle, I

would argue that if something is a conscious state, it must have a first

person experiential aspect, there must be something that it is "like" (in the Nagelian sense) to be in this state.33 The subjective, experiential

30 For intentionality in prelinguistic children, see, for example, Daniel N.

Stern, The Interpersonal World of the Infant: A View from Psychoanalysis and Developmental Psychology (New York: Basic Books, 1985), and Jerome Bruner, Child's Talk: Learning to Use Language (New York: W. W. Norton & Company, 1983). For a criticism of stimulus-response explanations of animal

behavior, as well as a defense of folk-psychological explanations as inelim inable and scientific, see Jerry Fodor, The Language of Thought (Cambridge:

Harvard University Press, 1975). 31 Sellars, "Phenomenalism," in Science, Perception and Reality, 97 n. 1. 32 The changing tide against behaviorism within empirical research has

been registered by Sellars's successors in various ways. Similarly to Sellars, Dennett is attracted by the promise of a possible reduction of the mental to the physical held out by stimulus-response explanations of behavior. How ever, unlike Sellars, he acknowledges the relative explanatory power and ir

reducibility of intentional explanations (explanations attributing beliefs and desires), even for nonverbal "systems" such as children and animals. He is able to combine his physicalism with the assertion of the irreducibility of in tentional explanation by jettisoning Sellars's scientific realism in favor of an instrumentalist interpretation of intentionality. See, for example, Daniel Den

nett, Content and Consciousness (New York: Routledge, 1993). More di rectly, although Brandom limits his mammoth analysis in Making it Explicit to linguistic intentionality, he concedes that this is only its "fanciest" form.

Unlike Sellars, he admits that animals and children have a simpler, nonlin

guistic intentionality and notes that it would be important to explain how the fancier, linguistic form emerges from the simpler, nonlinguistic ones. See

Robert B. Brandom, Making it Explicit: Reasoning, Representing, & Dis cursive Commitment (Cambridge: Harvard University Press, 1994), 7. It goes without saying that Husserl's analysis of prepredicative judgment ad dresses precisely this latter issue.

33 According to Nagel and Searle, this is an essential characteristic dis

tinguishing mental from physical states. See Thomas Nagel, "What is it like to be a bat?" in Mortal Questions (Cambridge: Cambridge University Press, 1979), and John Searle, The Rediscovery of the Mind (Cambridge: MIT Press, 1992). Husserl would surely share this view.

318 GAIL SOFFER

dimension is an essential feature of consciousness, it belongs to its

very definition as conscious. This does not mean that a conscious

state must be directed to an object or that a conscious state cannot be

subliminal: some conscious feelings may not be "of" anything, some

conscious states may be unaccompanied by attention or thematiza

tion. However, it does mean that there is always and necessarily (by

definition) an experiential dimension of a conscious state for the be

ing who is in that state, and that it is possible to give some account of

this experiential dimension. If this is impossible, then it is simply

empty and misleading to apply the term "conscious" to this type of

state. It is not enough to say, as does Sellars, that sensations are in

the mind in the same way that molecules are in a gas.34 A molecule is

a physical entity which is in a gas and causally affects its behavior in

dependently of any experience on the part of the gas. If a sense im

pression is a conscious rather than a physical state, it should have a

conscious mode of inherence and causality, which means that there is

something it is "like" to be in such a state.35

What then is the properly subjective, experiential dimension of

sensation, according to Sellars? What is it "like" to sense? Sellars

characterizes sensations as nonphysiological, conscious, and even

34 "These episodes are 'in' language-using animals as molecular impacts are 'in' gases, not as 'ghosts' are in 'machines'"; Sellars, Empiricism and the

Philosophy of Mind, 104. 351 cannot agree with Chalmers, who allows that there are two equally

valid concepts of consciousness, the phenomenal one (where conscious states are characterized by experience), and a "psychological" one (where conscious states are characterized functionally by the causal role they play in producing behavior). If there is something that has no experience (that is, if it is a zombie), nothing useful is added by calling its functional states "psy chological." Chalmers argues that many human conscious states are best un

derstood in this functionalist way, but I find his examples unconvincing. For

example, Chalmers holds that we usually understand learning functionally, as pure, experience-empty adaptation to the environment. To the contrary, it seems to me only the reductive distortions of cognitive science or artificial

intelligence would lead one to accept the very sad equation of human learn

ing with unconscious stimulus-response conditioning or, even worse, the al teration of behavior via direct physical intervention (for example, brain sur

gery). Precisely what distinguishes learning is the conscious, experienced apprehension and comprehension of the material on the part of the student. Otherwise we do not have learning but behavior modification via condition

ing, physical intervention, or changes in programming. For the opposing view of Chalmers, see David Chalmers, The Conscious Mind (Oxford: Ox ford University Press, 1996), 11 and following.

HUSSERL AND SELLARS ON THE GIVEN 319

"perceived." But he denies that they can be apperceived, noticed, or

observed.36 As with most theoretical entities, once one has learned

the theory (the language game), sensations can be directly known, in

the sense that it is possible noninferentially to report their presence.

For example, the sophisticated tie salesman, John, can noninferen

tially report that he is having a green sensation when he looks at the

blue tie in artificial light. However, in this case John is not seeing or

describing a green sensation. Rather, he is seeing qualitatively the

same thing as when he sees a green object, withholding his endorse

ment of what he sees, and giving a theoretical interpretation of his ex

perience in terms of the language game of sense impressions. Here "I

have a green sense impression" means "I am in the state that is ordi

narily caused by a green object under standard conditions, and this

state is formally analogous to the visible surface of the green object."

However, having a green sense impression is not "like" seeing a green

object or a green patch on the surface of an object. Sellars empha

sizes that a green sense impression is not itself green, a sense im

pression of a triangle is not itself triangular.37 Having a green sen

sation is not like seeing a patch of green, because then sensing would

be relational, intentional, cognitive and so would be in contradiction

with psychological nominalism. For Sellars, patches of green are seen

only by linguistically mature persons, and seeing the patch of green is

the result of sensation plus language and concepts. This is why he in

sists that the model for a green sensation is not seeing the patch of

green but the green surface itself. Thus the model for a sensation,

36 "Sense impressions are non-conceptual states of consciousness.

. . .

The phrase 'object of consciousness' is itself highly ambiguous but for the

moment, at least, I shall use it as roughly equivalent to 'noticed.' Like bodily sensations, visual impressions were construed as not only states but as, at least on occasion, objects of consciousness. Whatever Descartes himself may

have thought, there is nothing absurd in the idea that states of consciousness occur which are not apperceived, a fact which was appreciated by Leibniz.

More startling, and to many absurd, is the idea that there are broad classes of states of consciousness none of the members of which are apperceived. Star

tling or absurd, the idea is at least not obviously self-contradictory_In any case, I shall push it to the hilt"; Wilfrid Sellars, Science and Metaphysics: Variations on Kantian Themes (London: Routledge & Kegan Paul, 1968), 9 10.

37 Sellars, Empiricism and the Philosophy of Mind, 112-14.

320 GAIL SOFFER

supposedly something mental and conscious, turns out to be some

thing physical.38 But what does this analogy tell us about the experi ential dimension of sensation, about what it is like to have a sensa

tion? It would be absurd to say that having a green sensation is like

being a patch of green since there is no experiential dimension to be

ing an inanimate physical object. I would argue that interpreted strictly, Sellars's theory of sensa

tions methodologically excludes any answer to the question of what it

is like to have a sensation. The theory does not allow a specification of the experiential properties but only of the formal-logical relations

among the properties (that is, these formal-logical relations have to

match the formal-logical relations obtaining among the corresponding

models). Indeed, on Sellars's view, this is an essential feature and

merit of the functional definition of sensation, since the exclusion of

any reference to experience leaves open the possibility of an eventual

reduction of sensations to purely physical states. If it were essential

to sensations to be experiences, rather than merely to preserve the

same formal-logical relations as experience, then no physicalistic re

duction of sensation would be possible.39

38 "Here it is essential to note that the analogy is between sense impres sions and physical objects and not between sense impressions and percep tions of physical objects. Failure to appreciate this fact reinforces the temp tation to construe impressions as cognitive and conceptual which arises from the assimilation of the 'of-ness' of sensation to the 'of-ness' of thought. It is also essential to note that the analogy is a trans-category analogy, for it is an analogy between a state and a physical thing"; Sellars, "Phenomenalism," in Science, Perception and Reality, 93. 39

According to Sellars, the chief obstacle to a physicalistic reduction of the mental is that neurophysiological processes as currently understood do not display the same kind of logic as the manifest models of sensations. For

example, colored patches are continua, whereas neurons are discrete, so it is not clear what brain state would correspond to the infmitesimally small col ored patches of a visible colored surface. Sellars does not see the experien tial dimension of sensations as a problem for a physicalistic reduction. On his account, the experiential dimension belongs only to the manifest image, and so is not "really real." See Sellars, "Philosophy and the Scientific Image of Man," 35. In contrast to Sellars's physicalist version of scientific realism, in the Crisis and Ideas II, Husserl argues that the assertion of the reality of the world of science presupposes the assertion of the reality of the lifeworld, and that the absolutization of physical nature typical of naturalism and Car tesian-Galilean science is one of the most spiritually and morally perilous de

velopments in the history of European thought. See Husserl, The Crisis of the European Sciences, ??9h, 36, 55; Gail Soff er, "Phenomenology and Scien tific Realism."

HUSSERL AND SELLARS ON THE GIVEN 321

If we accept the view that experience is essential to conscious

states, then the same doubt about whether Sellarsian sensations count

as states of consciousness applies equally to Sellarsian thoughts. These are also characterized functionally as states formally analogous to acts of speech.40 There is nothing in this theoretical concept of a

thought that makes it necessary for thought to have an experiential di

mension for the being who has the thought. But thought for Sellars in

cludes perception, desire, and all other intentional states, and thought

together with sensation exhaust the sphere of consciousness. This

means that Sellars's entire theory of consciousness systematically ren

ders inessential what, according to thinkers such as Nagel, Searle, and

Husserl, is its most important distinguishing feature.41

Husserl also proposes an analysis of sensations, and the differ

ences between his account and that of Sellars are illuminating. First

and foremost, this account respects the crucial notion that conscious

experience has an experiential dimension. Thus Husserl emphasizes that sensations (Empfindungen) have a dual nature and can be con

ceptualized and studied in two different ways: either as stimulus-re

sponse reactions of the body to material events (for example, in phys

iology) or as the psychic "stuff worked over by psychological mechanisms of reproduction, association, and so forth (for example,

in psychology). Psychic sensations are the result of many processes

of synthesis but are "stuff" in the relative sense that they are at a stage of constitution that falls short of perception of fullfledged physical ob

jects. Psychic sensations are not usually noticed, but as states of con

sciousness they remain possible objects of notice. For example, we

could say that when a painter uses many different, widely ranging col

ors to depict a solid colored object, he is noticing and painting the sen

sations which would be sensed under the relevant environmental con

ditions. Sellars's claim that red sense impressions are not themselves

red (or any other color) but only have properties formally analogous

40 Sellars, "Philosophy and the Scientific Image of Man," in Science, Per

ception and Reality, 32-3; and Empiricism and the Philosophy of Mind, 105-8.

41 The tendency to eliminate the experiential dimension from the mental is particularly clear in the Sellars-Chisholm correspondence: "My claim is that the categories of intentionality are nothing more nor less than the meta

linguistic categories in terms of which we talk epistemically about overt

speech as they appear in the framework of thoughts construed on the model of overt speech"; Sellars and Chisholm, "Intentionality and the Mental," 522.

322 GAIL SOFFER

to the experiential qualities of things would apply at most to sensa

tions in the physiological sense, which are physical states of the

body.42

Even greater difficulties face the Sellarsian theory of sensation

taken as an account of prelinguistic consciousness. Minimally, this

would entail that without verbal language there could not be coherent

awareness or the attentive perception of individual objects. I have al

ready pointed out that this is not the prevailing view in empirical re

search involving prelinguistic children and intelligent animals. In con

cluding this section, I will raise three further doubts about this

position: (1) If we generally explain certain behavior of linguistic be

ings by appealing to their attentive perception of objects, why should

we not explain the very same behavior in prelinguistic, physiologi

cally similar beings in the same way? (2) How can verbal language

confer attentive perceptions, inferential processes, and so forth, upon

a being who does not already possess them? (3) How can we plausi

bly explain the way children in fact acquire language without attribut

ing to them the attentive perception of objects and certain other rudi

mentary forms of rationality?

It is undeniable that the acquisition of language brings about cru

cial changes in behavior and experience. Sellars is quite right to argue

against classical empiricism that language does not merely supervene

upon a solipsistically completed logical space of meaning. Similarly to classical empiricism, Husserl fails to elucidate the role of verbal

language at the lower levels of object constitution. It is also clearly correct to insist with Wittgenstein and Sellars, and contrary to classi

cal empiricism, that in the case of many (j)'s, first we learn to speak about (?), and only later acquire the concept of c(). It is also true that

the meaning of many words and phrases cannot be analyzed except in

terms of their role in language games.43 However, the acquisition of

conventional verbal speech does not bring about the universal trans

formation in behavior and experience one would expect with a first

acquisition of attentive perception of objects. Thus, Sellars goes too

far when he holds that there is no logical space of meaning prior to

verbal language. Some prelinguistic behavior is of the same type as

42 For an account of sensation, see Edmund Husserl, Ideen zu einer reinen Ph?nomenologie und ph?nomenologischen Philosophie, Drittes

Buch, ed. Marly Biemel, Husserliana, vol. 5 (The Hague: Martinus Nijhoff, 1952), 14-16.

43 Along the lines of Sellars's view that "means" functions as in "Rot

means red in German." In this case, "x means y" means that x plays the same

role in the language as y.

HUSSERL AND SELLARS ON THE GIVEN 323

the nonverbal behavior of linguistic beings that we generally explain

by appealing to object perception and various forms of cognition. For

example, prelinguistic infants smile at the approach of their mothers

(but not at strangers), startle and cry at loud noises, stare at objects, follow them with their gazes, point at them, reach for them, bring about and prepare for future events. When linguistically mature per sons do the same, we say that they see the objects they are looking at,

they recognize them, they have desires, they anticipate, infer, and so

forth. If there really is a dramatic difference in the logical coherence

of verbal and preverbal experience, should there not be a correspond

ing difference in behavior? Why should the very same behavior have an intentional explanation in the one case but not in the other?44

Further, if there is no object perception behind seemingly intelli

gent prelinguistic behavior, then what is its explanation? If we at

tempt a stimulus-response explanation, we cannot appeal to the indi

vidual's perceiving the stimuli for, by hypothesis, the individual does

not notice anything, not even his own sensory states. We cannot say, for example, that a newborn infant settles more quickly when picked

up by his own mother because he sees his mother's face, feels her

arms, recognizes that this is his mother (for example, the same face

and touch he usually perceives when being held). When he cries be

fore eating, we cannot say that he feels pangs of hunger. Similarly, when he fixes his eyes on a moving object and follows it with his gaze, we cannot say that he is focusing his eyes and turning his head be

cause he sees the object and wants to continue seeing it.

One possibility would be to give a purely physical account of the

causation of behavior, as in the case of reflexes. However, an obvious

objection to the physicalist account of prelinguistic differential behav

ior is that some of the behavior of prelinguistic infants and animals, unlike reflexes, evolves very rapidly and takes on forms highly

adapted to the environment, while the underlying physical structure of

the brain and body is relatively fixed. Thus, it is unclear how a purely

44 In his correspondence with Chisholm, Sellars maintains that attribu tions of intentionality to animals are always qualified to such an extent that they are not really attributions of intentionality in the same sense as we at tribute to humans. See Sellars and Chisholm, "Intentionality and the Mental," 527. However, even if we concede that intentionality is affected by language, it is not the case that we never attribute the same intentionality to prelinguis tic beings as to mature persons. For example, who feels tempted to say of an infant attentively focusing upon and clamoring for an object, "He sees it and

wants to touch it, but he doesn't really see it and doesn't really want to touch it?"

324 GAIL SOFFER

physical mechanism could give rise to adaptive, learned behavior.

Further, if a sufficiently adaptive neurophysiological mechanism

could be found, then it could be used to explain the intelligent behav

ior of linguistic beings as well, so the differentiation between linguis

tic and nonlinguistic beings would again be lost.45

A second question concerns how the acquisition of verbal speech

could confer perception and attentive awareness upon a being who

did not already possess it, thereby transforming experientially incho

ate sensory states into attentive perceptions of things. Indeed, at the

stage where there is only sensation, we should not even speak about

the acquisition of language but about the acquisition of the physical

behavior corresponding to language. On Sellars's model, we must as

sume that at the start the production of wordlike sounds is unaccom

panied even by hearing in the usual sense since hearing involves no

ticing sounds, reidentifying them over time, recognizing that the

phoneme heard today in a high pitched tone is the same phoneme as

the one said in a low voice yesterday, and other cognitive functions

which, according to Sellars, presuppose language. Thus the infant's

first word production is not even at the level of a parrot, which we

usually think of as hearing sounds without understanding them (that

is, in a way similar to the way adults hear speech in an unknown for

eign language). Rather, Sellars's infant seems closer to a tape re

corder or a speech synthesizer, producing sounds with no or only in

choate auditory experience. Yet if language acquisition begins in this

way, how can language get intentionality off the ground? How can

even the most rule-governed sound production without coherent, at

tentive hearing move the infant across the abyss to sound production

with coherent, attentive hearing? Following Dennett's evolutionary

45 Dennett, for example, defends physicalism by arguing that the brain

itself evolves, so that new behavioral response mechanisms can be as a re

sult new or altered afferent-efferent connections, without appealing to con

sciousness or its causality. See Daniel Dennett, Content and Consciousness

and Consciousness Explained (Boston: Little, Brown & Co., 1991). A related

view is presented by Churchland. See Paul Churchland, A Neurocomputa tional Perspective (Cambridge: MIT Press, 1992). However, Dennett and

Churchland offer more of a research program than a convincing argument.

Further, their approaches drop Sellars's essential distinction between pre and postlinguistic consciousness and his insistence that intentionality re

quires language.

HUSSERL AND SELLARS ON THE GIVEN 325

model of the brain, we might argue that the acquisition of linguistic

physical behavior changes the neural wiring of the brain, and this in

turn gives rise to attentive awareness. However, this claim rests not

only on a research program but on a moribund research program

since no empirical research supports a distinction between attentively aware brains and inattentively aware brains in correlation with the de

velopment of the linguistic centers of the brain.

Similar problems arise for the explanation of language acquisition itself. Language acquisition in humans is a complex process which

contemporary researchers have found impossible to explain without

appealing to many cognitive skills, such as attention direction, pattern

recognition, and predictive inference. In order to learn how to use a

word in socially instituted language games, one has to hear it spoken.

Phonemes have to be reidentified over a vast range of different

pitches, accents, voices, and environmental conditions. Children have

to learn not just to produce and recognize regularly patterned sounds

but also that specific sounds refer to specific objects and that specific sounds can be used to bring about specific events in the world. The

ability to understand that verbal sounds refer to things in the world

presupposes the ability to understand the referentiality of nonverbal

gestures, such as points and attentive directing of the gaze.46 If lan

guage is a game, learning to play a game itself requires inferential

skills, such as the ability to predict the next appropriate move, the

move that will bring about the desired result.47 Children require a pre

linguistic ability to recognize a situation as a certain stage of the game and to infer which move should be made next, in order to be able to

learn the game of language at all. There is no empirical evidence that

children could acquire language unless they already possessed these

forms of perception and rational skills.48

It seems to me that these considerations raise significant doubts

about Sellars's position that prior to language there are only raw

46 See Bruner, "The Growth of Reference," in Child's Talk: Learning to Use Language (New York: W. W. Norton & Company, 1983). 47 For a discussion of the inferential skills required for game-playing, and their relevance to language acquisition, see Bruner, "Play, Games, and Lan

guage," in Child's Talk: Learning to Use Language. 48 Whether a computer could simulate or even "acquire" language skills without intentionality is not relevant in this context. However a computer might do it, the point is that this is not how children do it.

326 GAIL SOFFER

sensory states without attention, recognition, inference, or

intentionality.49 However, if there is any prelinguistic intentionality at

all, then language cannot be the essence of intentionality. The prob

lem with Sellars's interpretation of the linguistic approach to inten

tionality is not merely that it misconstrues the experience of infants

and animals, or that it overlooks the prepredicative level of adult ex

perience. Rather, the deeper problem is that it tends to obscure the

fundamental experiential dimension of intentionality. In the conclud

ing section, I will trace the roots of this flight from experience to a

flawed understanding of intersubjectivity typical of postpositivistic analytic philosophy.

IV

If positivistic analytic philosophy was dominated by Cartesian

anxiety, the anxiety for absolute certainty about the world, a certain

portion of contemporary analytic philosophy of mind remains under

the sway of another anxiety. For we must wonder, what really is be

hind the ubiquitous flight from consciousness, from subjectivity, from

the experiential sphere? Why is subjectivity reduced if not to neuro

physiology or behavior, then to language, but in any case to something

third person at any cost?

Here we can distinguish a wide range of overlapping motives

coming from different directions and influencing different strands of

thought. Within empirical psychology, the turn to behaviorism arose

as a reaction to the controversies over introspective descriptions of

experience.50 In this case the flight from subjectivity is a flight to the

relative objectivity and publicity of the third person domain. Searle,

on the other hand, explains the popularity of physicalism in the con

temporary philosophy of mind in terms of the fear of Cartesian dual

ism and its well-known paradoxes, especially as outlined by Ryle.51 In

thinkers such as Heidegger and Gadamer, the turn away from con

sciousness to language represents a practical-historical critique of

49 For a more detailed critique of this position and a discussion of prelin guistic conception formation, see my "Language and the Formation of Gen

eral Concepts," in HusserVs Logical Investigations, ed. Daniel O. Dahlstrom

(Dordrecht: Kluwer, 2003), 37-56. 50 L. S. Vygotsky, Mind in Society: The Development of Higher Psycho

logical Processes (Cambridge: Harvard University Press, 1979), 3-4. 51

Searle, The Rediscovery of the Mind, 2 and following.

HUSSERL AND SELLARS ON THE GIVEN 327

Cartesianism, emphasizing that thought is neither purely theoretical

reflection nor a solipsistic first beginning but a historically and so

cially conditioned engagement with the world. A similar stress on the

social, practical, and historical nature of the mind emerges from the

Marxist tradition, resulting in a linguistic turn in figures such as

Habermas and Vygotsky. Clearly, all these various motives come to

gether in the thought of Sellars.

However, I would suggest that the idea of the nonobjectivity and

even unfathomability of the subjective receives an additional existen

tial impetus from a certain type of encounter with the other. This en

counter has been described well by Sartre, with his usual exaggera

tions and theatricality, in his discussion of the Look in Being and

Nothingness.,52 According to Sartre, in the encounter with the other as

Look, the other's subjectivity as subjectivity, the other's experience as

experience, appears completely inaccessible. The other appears to be

a drainhole in my world sucking up everything, especially my own ob

jective being. For Sartre, the other as Look is the source of a particu

lar kind of anxiety which we might call "alterity anxiety." It is difficult

to avoid the suspicion that a certain measure of alterity anxiety rein

forces some of the contemporary flight from the subjective dimension

of consciousness, in its various manifestations.

According to Sartre's analysis, the experience of the other as

Look gives rise to a number of typical strategies to get the genie of the

other's subjectivity back into the bottle, to stop the hemorrhaging of

the world. The most straightforward such strategy is the objectifica

tion of the other. The widespread view in the philosophy of mind that

intentionality is primarily a theory for the explanation of behavior fits

well under the Sartrean rubric of objectification of the other. The idea

that the other's mental life is a theoretical hypothesis ensures that the

subjectivity of the other is externalized, knowable, and dependent

upon the subjectivity of the self for its reality (if real at all). However, as already seen by Sartre, such strategies generate their own dialectic,

and this one is no exception. For applied to the self, it results in the

paradoxical conclusion that one's own intentionality is a theoretical

hypothesis, thus essentially transforming, not to say eliminating, the

experiential dimension.

52 Jean-Paul Sartre, "The Look," in Being and Nothingness, trans. Hazel Barnes (New York: Washington Square Press, 1956).

328 GAIL SOFFER

The conception of intentionality as a theory for the explanation of behavior plays an important role within the philosophy of mind in

general, and in the thought of Sellars in particular. This conception reaches its most extreme form with Dennett's notion of the inten

tional stance, according to which a system (including a computer sys

tem) is intentional if and only if its behavior can be successfully ex

plained and predicted by attributing it thoughts, desires, beliefs, and

so forth. Sellars's version of this position differs from Dennett's in

certain important respects. In particular, Sellars distinguishes be

tween intentionality itself or the having of mental states, on the one

hand, and the conceptual framework of intentionality or the attribut

ing of mental states, on the other. For Sellars, it is only the concep tual framework of intentionality (folk psychology) that is a theory for the explanation of behavior (especially verbal behavior).

Nonetheless, I believe the Sellarsian conception is seriously flawed and fails to do justice to significant insights regarding intersub

jectivity found in the phenomenological tradition. This distorted ap

proach to intersubjectivity in turn supports the flight from experience

typical of the linguistic/functionalist conception of consciousness. In

particular, Sellars claims to solve the problem of other minds by

building the fact that behavior is evidence for mental states ("inner

episodes") into the very logic of the concept of a mental state. To the

contrary, I will argue that a correct approach to intersubjectivity shows that the problem of other minds cannot be solved, and a rela

tion to behavior does not and should not be made to belong to the

concept of a mental state.

In order to support my position in detail, it will be useful briefly to review the Jones myth from Empiricism and the Philosophy of

Mind, which contains Sellars's account of intersubjectivity and "solu

tion" of the problem of other minds. In this myth, Sellars opposes the

traditional empiricist genetic account, according to which we first

form the concept of an inner episode by reflecting upon our own "im

mediately given" inner episodes. On the classical account, we have

reflective access to our thoughts independently of initiation into the

socio-linguistic sphere, whose role is reduced to fixing the conven

tional sounds and marks to be attached to our solipsistically consti

tuted thoughts. Part and parcel of this view is the classical theory of

meaning, "S means that p" means "S expresses the thought that p." Since knowledge of inner episodes is conceived as essentially solip

sistic, this view gives rise in turn to the problem of other minds. We

know our own inner episodes by solipsistic introspection. How are

HUSSERL AND SELLARS ON THE GIVEN 329

we to know the inner episodes of others, or that others have inner epi

sodes at all?

All of this, argues Sellars, belongs squarely within the Myth of the

Given. Sellars opposes this view by constructing his own mythical ac

count of the origin of the language and conceptual framework of men

tal states ("inner episodes"). According to Sellars's account: (1) se

mantic language occurs prior to the concept of an inner episode, so

there is a concept of meaning which does not involve reference to

thoughts; (2) the concept of an inner episode derives from and presup

poses language and the social sphere; and therefore (3) there is a es

sential link between inner episodes and public behavior, so the prob

lem of other minds does not arise.

The Jones myth is well known, so a few brief remarks will suffice.

Sellars imagines a population possessing a Rylean, behavioristic lan

guage with terms referring to physical objects and events, including

behavior and overt speech acts but no mental terms. At the second

stage, Jones introduces the concept of a thought as a theoretical hy

pothesis for the explanation of behavior. He conceives of thoughts as

formally analogous to acts of silent speech, and he theorizes that they cause intelligent behavior and sometimes result in acts of overt

speech.53 Jones then teaches the other Ryleans how to use this ex

planatory strategy, so that at the third stage each person regularly draws inferences from the behavior of others to their thoughts and

similarly draws inferences about his own thoughts by observing his

own behavior. At the fourth stage, the Joneseans can noninferentially

report their own thoughts, perceptions, and other intentional states

without observing their own behavior. At the fifth stage they arrive at

the notion of sense impressions as part of a causal theory of percep

tion, in the manner described in the previous section.54

Now no one, including Sellars, would be tempted to see the Jones

myth as a plausible historical or individual developmental account.

He is not claiming that first we have semantic language without men

tal language and only later develop the language of mental states. He

is also not claiming that intentionality itself is or presupposes a

53 "Clearly the episodes in which we are interested are not shifting be

havioral properties; they are connected with such shifts, but the connection is synthetic, as in the connection of molecular motion with the shifting pro

pulsive propensities of a volume of gas"; Sellars, Science and Metaphysics, 73 n. 29.

54 Sellars, Empiricism and the Philosophy of Mind, 91-3.

330 GAIL SOFFER

theoretical explanation of behavior, since in the myth the Ryleans

have thoughts without knowing that they have thoughts even before

they develop the theoretical framework of thoughts.55 However, he

does claim that:

(1) it is at least possible that people could live together and develop a semantic language without attributing intentionality to others or to

themselves;56 and

(2) the inferential, theoretical attribution of intentionality (that is, to ex

plain the behavior of others) precedes the noninferential attribution of

intentionality (that is, to oneself).57

We should be clear about the importance of each of these two

points within Sellars's position as a whole. The first point supports

his claim that the concept of meaning does not require reference to

mental states but can be elucidated as a metalinguistic concept.58 The

second point insures that even where intentional states can be re

ported or "introspected," they retain an essential relation to outward

behavior and so are public. As Sellars puts it, the fact that behavior is

evidence for inner episodes is built into the very logic of the concept

of an inner episode.59

However, I will argue that both (1) and (2) are false and based on an inadequate analysis of intersubjectivity. Moreover, it is not possi

ble first to encounter the other as intentional via inference or theoret

ical hypothesis; and Sellars does not show that it possible for a popu

lation to develop a Rylean language or even to live together in society

without already attributing intentionality to others.

55 "They think, but they don't know that they think. Their use of lan

guage is meaningful because it is the expression of thoughts, but they don't know that it is the expression of thoughts; that is to say, they don't know that overt speech is the culmination of inner episodes of a kind which we con ceive of as thoughts"; Sellars and Chisholm, "Intentionality and the Mental," 526.

56 "I have argued that it is in principle possible to conceive of the charac teristic forms of semantical discourse being used by a people who have not

yet arrived at the idea that there are such things as thoughts"; Ibid. 57 There seems to be some wavering on Sellars's part between holding

that inferential attribution of intentionality does precede its noninferential

attribution, at least as a logical precondition, and holding that it could pre cede it in time. My argument is not affected by this ambiguity.

58 urYhe argument presumes that the metalinguistic vocabulary in which we talk about linguistic episodes can be analysed in terms which do not pre suppose the framework of mental acts"; Sellars and Chisholm, "Intentionality and the Mental," 522.

59 Sellars, Empiricism and the Philosophy of Mind, 107.

HUSSERL AND SELLARS ON THE GIVEN 331

Many philosophers in t